Count of unique Subsequences of given String with lengths in range [0, N]

Given a string S of length N, the task is to find the number of unique subsequences of the string for each length from 0 to N.

Note: The uppercase letters and lowercase letters are considered different and the result may be large so print it modulo 1000000007.

Examples:

Input: S = “ababd”
Output: 
Number of unique subsequences of length 0 is 1
Number of unique subsequences of length 1 is 3
Number of unique subsequences of length 2 is 6
Number of unique subsequences of length 3 is 8
Number of unique subsequences of length 4 is 5
Number of unique subsequences of length 5 is 1

Explanation: 0 length subsequences are 1-> {}
1 length subsequences are 3 -> a, b, d
2 length subsequences are 6 -> ab, aa, ad, ba, bd, bb
3 length subsequences are 8 -> aab, aad, abb, abd, bab, aba, bbd, bad
4 length subsequences are 5 -> aabd, abab, abad, babd, abbd
5 length subsequences are 1 -> ababd

Input: GeeksForGeeks
Output: 
Number of unique subsequences of length 0 is 1
Number of unique subsequences of length 1 is 7
Number of unique subsequences of length 2 is 43
Number of unique subsequences of length 3 is 163
Number of unique subsequences of length 4 is 402
Number of unique subsequences of length 5 is 703
Number of unique subsequences of length 6 is 917
Number of unique subsequences of length 7 is 918
Number of unique subsequences of length 8 is 711
Number of unique subsequences of length 9 is 421
Number of unique subsequences of length 10 is 185
Number of unique subsequences of length 11 is 57
Number of unique subsequences of length 12 is 11
Number of unique subsequences of length 13 is 1

Naive Approach: The basic approach to solve the problem is as follows:

Generate every possible subsequence and store it in a set to get unique results. Then traverse each of the subsequence from that set and count it on the basis of their lengths.

Follow the steps to solve the problem:

• Use recursion to generate each subsequence and store it in a set to get unique occurrences.
• Use a map to store the count of subsequences for each possible length.
• Traverse each subsequence, find the length of the subsequence and update the count of the appropriate group of subsequences.
• Traverse map and print both its keys (length of subsequence) & value (count of subsequence).

Below is the implementation for the above approach:

C++14

// C++ code to implement the above approach.

#include <bits/stdc++.h>
using namespace std;
#define MAX 1000000007
typedef long long ll;
unordered_set<string> sn;

// Function to find subsequences
void subsequences(char s[], char op[],
                  ll i, ll j)
{
    if (s[i] == '\0') {
        op[j] = '\0';
        sn.insert(op);
        return;
    }
    else {
        op[j] = s[i];
        subsequences(s, op, i + 1, j + 1);
        subsequences(s, op, i + 1, j);
        return;
    }
}

map<ll, ll> getCount()
{
    map<ll, ll> freq;

    for (auto x : sn) {
        freq[x.length()]
            = (freq[x.length()] + 1) % MAX;
    }

    return freq;
}

// Function to print the answer
void printSubsequences(string& S)
{
    ll m = S.length();
    char op[m + 1];

    subsequences(&S[0], &op[0], 0, 0);
    map<ll, ll> ans = getCount();

    for (auto& x : ans)
        cout << "Number of unique subsequences of length "
             << x.first << " is " << x.second << endl;
}

// Driver Code
int main()
{
    string S = "ababd";

    // Function call
    printSubsequences(S);
    return 0;
}

Java

// Java code to implement the approach
import java.io.*;
import java.util.*;

class GFG {
    static int MAX = 1000000007;
    static HashSet<String> sn;
    // Function to find subsequences
    static void subsequences(String s, String op, int i)
    {
        if (i == s.length()) {
            sn.add(new String(op));
            return;
        }

        else {
            subsequences(s, op + s.charAt(i), i + 1);
            subsequences(s, op, i + 1);
            return;
        }
    }
    static HashMap<Long, Long> getCount()
    {
        HashMap<Long, Long> freq = new HashMap<>();

        for (String x : sn) {
            long len = x.length();
            freq.put(len, (freq.getOrDefault(len, 0L) + 1)
                              % MAX);
        }

        return freq;
    }

    // Function to print the answer
    static void printSubsequences(String S)
    {
        int m = S.length();
        String op = "";
        sn = new HashSet<>();
        subsequences(S, op, 0);
        HashMap<Long, Long> ans = getCount();

        for (Long x : ans.keySet())
            System.out.println(
                "Number of unique subsequences of length "
                + x + " is " + ans.get(x));
    }
    public static void main(String[] args)
    {
        String S = "ababd";

        // Function call
        printSubsequences(S);
    }
}
// This code is contributed by Karandeep1234

Python3

# Python3 code to implement the above approach.
MAX = 1000000007
sn = set()

# Function to find subsequences
def subsequences(s, op, i):
    global sn

    if (i == len(s)):
        sn.add("".join(op))

    else:
        subsequences(s, op + s[i], i + 1)
        subsequences(s, op, i + 1)

# Function to find count of subsequences
# with a particular length
def getCount():
    freq = dict()
    for x in sn:
        if len(x) in freq:
            freq[len(x)] = (freq[len(x)] + 1) % MAX
        else:
            freq[len(x)] = 1 % MAX
    return freq

# Function to print the answer
def printSubsequences(s):
    global op
    m = len(s)
    op = ""

    subsequences(s, op, 0)
    ans = getCount()
    for x in sorted(ans):
        print("Number of unique subsequences of length", x, "is", ans[x])

# Driver Code
s = "ababd"

# Function call
printSubsequences(s)

# This code is contributed by phasing17

C#

// C# code to implement the approach

using System;
using System.Collections.Generic;

class GFG {
    static int MAX = 1000000007;
    static HashSet<string> sn;

    // Function to find subsequences
    static void subsequences(string s, string op, int i)
    {
        if (i == s.Length) {
            sn.Add(new string(op));
            return;
        }

        else {
            subsequences(s, op + s[i], i + 1);
            subsequences(s, op, i + 1);
            return;
        }
    }

    // Function to build the dictionary of counts
    static Dictionary<long, long> getCount()
    {
        Dictionary<long, long> freq
            = new Dictionary<long, long>();

        foreach(string x in sn)
        {
            long len = x.Length;
            if (freq.ContainsKey(len))
                freq[len] = (freq[len] + 1) % MAX;
            else
                freq[len] = 1L;
        }

        return freq;
    }

    // Function to print the answer
    static void printSubsequences(string S)
    {
        int m = S.Length;
        string op = "";
        sn = new HashSet<string>();
        subsequences(S, op, 0);
        Dictionary<long, long> ans = getCount();

        List<long> keyList = new List<long>(ans.Keys);
        keyList.Sort();

        foreach(var key in keyList) Console.WriteLine(
            "Number of unique subsequences of length " + key
            + " is " + ans[key]);
    }

    // Driver code
    public static void Main(string[] args)
    {
        string S = "ababd";

        // Function call
        printSubsequences(S);
    }
}
// This code is contributed by phasing17

Javascript

// JavaScript code to implement the above approach.

let MAX = 1000000007;
let sn = new Set();

// Function to find subsequences
function subsequences(s, op, i)
{

    if (i == s.length)
        sn.add(op);

    else
    {
        subsequences(s, op + s[i], i + 1);
        subsequences(s, op, i + 1);
    }
}

// Function to find count of subsequences
// with a particular length
function getCount()
{
    let freq = {};
    for (let x of sn)
    {
        if (freq.hasOwnProperty(x.length))
            freq[x.length] = (freq[x.length] + 1) % MAX;
        else
            freq[x.length] = 1 % MAX;
    }
    return freq;

}

// Function to print the answer
function printSubsequences(s)
{
    let m = s.length;
    let op = "";

    subsequences(s, op, 0);
    let ans = getCount();
    let keys = Object.keys(ans);
    keys.sort();
    for (let x of keys)
        console.log("Number of unique subsequences of length", x, "is", ans[x]);
}


// Driver Code
let s = "ababd";

// Function call
printSubsequences(s);

// This code is contributed by phasing17

Output

Number of unique subsequences of length 0 is 1
Number of unique subsequences of length 1 is 3
Number of unique subsequences of length 2 is 6
Number of unique subsequences of length 3 is 8
Number of unique subsequences of length 4 is 5
Number of unique subsequences of length 5 is 1

Time Complexity: O(2^N)
Auxiliary Space: O(N)

Efficient Approach: The idea to solve the problem using dynamic programming is based on the following observations:

Observations:

Consider a character at i^th position to be the end character of the subsequence with length j. 
Let the total possible ways be denoted as f(i, j).

This value depends on the values of f(i-1, j) and f(i-1, j-1), i.e. it is the summation of f(i-1, j) and f(i-1, j-1). 
So f(i, j) = f(i-1, j) + f(i-1, j-1)

But if the i^th character has occurred earlier in an index k, then in the above case, value of f(k-1, j-1) is being considered twice: 

• once for f(k, j) and 
• next for f(i, j) [as f(k-1, j-1) is a part of f(i-1, j-1) and f(k, j) is part of f(i-1, j)] 

and both the time the resulting subsequences are same because the k^th and i^th character are same.
So, in that case, we need to consider the value only once. Therefore f(i, j) = f(i-1, j) + f(i-1, j-1) – f(k-1, j-1)

So there are two cases:

• f(i, j) = f(i-1, j) + f(i-1, j-1) when ith character does not occur earlier
• f(i, j) = f(i-1, j) + f(i-1, j-1) – f(k-1, j-1) when i^th character occurs earlier at k^th index. 

Follow the below steps to solve the problem:

• Create a 2-dimensional array dp[][] where dp[i][j] represents the number of unique subsequences of S until i-th element in string and subsequences are of length j.
• Create an array (say last[]) to store the previous occurrence of a character.
• Use the transition function shown above to calculate the value of dp[i][j]. 
• Base cases are dp[0][0]=1 and dp[i][j]=0 for every j>i.

Below is the implementation of the above approach:

C++14

// C++ code to implement the approach

#include <bits/stdc++.h>
using namespace std;
typedef long long ll;
#define mod 1000000007

// Function to find subsequences
void findSubsequences(string& s)
{
    int n = s.length();
    ll dp[n + 2][n + 2];
    memset(dp, 0, sizeof(dp));
    vector<ll> last(256, -1);

    dp[0][0] = 1;

    for (int i = 0; i < n + 1; i++) {
        for (int j = i + 1; j < n + 1; j++)
            dp[i][j] = 0;
    }

    // Loop to implement the dp transition
    for (int i = 1; i <= n; i++) {
        for (int j = 0; j < n + 1; j++) {
            dp[i][j] = (dp[i][j]
                        + dp[i - 1][j])
                       % mod;
            if (j >= 1) {
                dp[i][j]
                    = (dp[i][j]
                       + dp[i - 1][j - 1])
                      % mod;
                if (last[s[i - 1]] != -1) {
                    dp[i][j]
                        = (dp[i][j]
                           - dp[last[s[i - 1]]][j - 1])
                          % mod;
                }
            }
        }
        last[s[i - 1]] = (i - 1) % mod;
    }

    cout << "Number of unique subsequences of length 0 is 1"
         << endl;

    for (int i = 1; i <= n; i++)
        cout << "Number of unique subsequences of length "
             << i << " is " << dp[n][i] << endl;
}

// Driver code
int main()
{
    string S = "ababd";

    // Function call
    findSubsequences(S);
    return 0;
}

Java

// Java program for the above approach
import java.io.*;
import java.util.*;

class GFG
{
static int mod = 1000000007;

// Function to find subsequences
static void findSubsequences(String s)
{
    int n = s.length();
    int[][] dp = new int[n + 2][n + 2];
     for (int i = 0; i < n + 2; i++) {
        for (int j = 0; j < n + 2; j++)
            dp[i][j] = 0;
    }
    int[] last = new int[256];
    for (int i = 0; i < 256; i++) {
        last[i] = -1;
    }

    dp[0][0] = 1;

    for (int i = 0; i < n + 1; i++) {
        for (int j = i + 1; j < n + 1; j++)
            dp[i][j] = 0;
    }

    // Loop to implement the dp transition
    for (int i = 1; i <= n; i++) {
        for (int j = 0; j < n + 1; j++) {
            dp[i][j] = (dp[i][j]
                        + dp[i - 1][j])
                       % mod;
            if (j >= 1) {
                dp[i][j]
                    = (dp[i][j]
                       + dp[i - 1][j - 1])
                      % mod;
                if (last[s.charAt(i - 1)] != -1) {
                    dp[i][j]
                        = (dp[i][j]
                           - dp[last[s.charAt(i - 1)]][j - 1])
                          % mod;
                }
            }
        }
        last[s.charAt(i - 1)] = (i - 1) % mod;
    }

    System.out.println( "Number of unique subsequences of length 0 is 1");

    for (int i = 1; i <= n; i++)
        System.out.println("Number of unique subsequences of length "
                        + i + " is " + dp[n][i]);
}

// Driver Code
public static void main(String args[])
{
    String S = "ababd";

    // Function call
    findSubsequences(S);
}
}

// This code is contributed by sanjoy_62.

Python3

## Python program for the above approach:
mod = 1000000007

## Function to find subsequences
def findSubsequences(s):
    n = len(s);
    dp = [];
    for i in range(0, n+2):
        dp.append([0]*(n+2))

    last = [-1]*256

    dp[0][0] = 1;

    ## Loop to implement the dp transition
    for i in range(1, n+1):
        for j in range(0, n+1):
            dp[i][j] = (dp[i][j] + dp[i - 1][j]) % mod;
            if (j >= 1):
                dp[i][j] = (dp[i][j] + dp[i - 1][j - 1]) % mod;
                if (last[ord(s[i - 1])] != -1):
                    dp[i][j] = (dp[i][j] - dp[last[ord(s[i - 1])]][j - 1]) % mod
        last[ord(s[i - 1])] = (i - 1) % mod

    print("Number of unique subsequences of length 0 is 1")

    for i in range(1, n+1):
        print("Number of unique subsequences of length", i, "is", dp[n][i])

## Driver code
if __name__=='__main__':

    S = "ababd";

    ## Function call
    findSubsequences(S)

    # This code is contributed by subhamgoyal2014.

C#

// C# program for the above approach
using System;

class GFG
{
  static int mod = 1000000007;

  // Function to find subsequences
  static void findSubsequences(String s)
  {
    int n = s.Length;
    int[,] dp = new int[n + 2, n + 2];
    for (int i = 0; i < n + 2; i++)
    {
      for (int j = 0; j < n + 2; j++)
        dp[i, j] = 0;
    }
    int[] last = new int[256];
    for (int i = 0; i < 256; i++)
    {
      last[i] = -1;
    }

    dp[0, 0] = 1;

    for (int i = 0; i < n + 1; i++)
    {
      for (int j = i + 1; j < n + 1; j++)
        dp[i, j] = 0;
    }

    // Loop to implement the dp transition
    for (int i = 1; i <= n; i++)
    {
      for (int j = 0; j < n + 1; j++)
      {
        dp[i, j] = (dp[i, j]
                    + dp[i - 1, j])
          % mod;
        if (j >= 1)
        {
          dp[i, j]
            = (dp[i, j]
               + dp[i - 1, j - 1])
            % mod;
          if (last[s[i - 1]] != -1)
          {
            dp[i, j]
              = (dp[i, j]
                 - dp[last[s[i - 1]], j - 1])
              % mod;
          }
        }
      }
      last[s[i - 1]] = (i - 1) % mod;
    }

    Console.WriteLine("Number of unique subsequences of length 0 is 1");

    for (int i = 1; i <= n; i++)
      Console.WriteLine("Number of unique subsequences of length "
                        + i + " is " + dp[n, i]);
  }

  // Driver Code
  public static void Main()
  {
    String S = "ababd";

    // Function call
    findSubsequences(S);
  }
}

// This code is contributed by saurabh_jaiswal.

Javascript

// JavaScript program for the above approach
const mod = 1000000007;

// Function to find subsequences
function findSubsequences(s)
{
    var n = s.length;
    s = s.split("");

    // mapping the string s to an array
    // corresponding to the ascii code
    // of each letter in s
    for (var i = 0; i < n; i++)
        s[i] = s[i].charCodeAt(0);
    var dp = [];
    for (var i = 0; i < n + 2; i++)
        dp.push(new Array(n + 2));
     for (var i = 0; i < n + 2; i++) {
        for (var j = 0; j < n + 2; j++)
            dp[i][j] = 0;
    }
    var last = new Array(256).fill(-1);


    dp[0][0] = 1;

    for (var i = 0; i < n + 1; i++) {
        for (var j = i + 1; j < n + 1; j++)
            dp[i][j] = 0;
    }

    // Loop to implement the dp transition
    for (var i = 1; i <= n; i++) {
        for (var j = 0; j < n + 1; j++) {
            dp[i][j] = (dp[i][j]
                        + dp[i - 1][j])
                       % mod;
            if (j >= 1) {
                dp[i][j]
                    = (dp[i][j]
                       + dp[i - 1][j - 1])
                      % mod;
                if (last[s[i - 1]] != -1) {
                    dp[i][j]
                        = (dp[i][j]
                           - dp[last[s[i - 1]]][j - 1])
                          % mod;
                }
            }
        }
        last[s[i - 1]] = (i - 1) % mod;
    }

    console.log( "Number of unique subsequences of length 0 is 1");

    for (var i = 1; i <= n; i++)
        console.log("Number of unique subsequences of length "
                        + i + " is " + dp[n][i]);
}

// Driver Code
var S = "ababd";

// Function call
findSubsequences(S);

// This code is contributed by phasing17

Output

Number of unique subsequences of length 0 is 1
Number of unique subsequences of length 1 is 3
Number of unique subsequences of length 2 is 6
Number of unique subsequences of length 3 is 8
Number of unique subsequences of length 4 is 5
Number of unique subsequences of length 5 is 1

Time Complexity: O(N²)
Auxiliary Space: O(N)